Anthropic Suspends AI Models After US Export Control Order
"The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees," the company stated, clarifying that all other models in its lineup remain unaffected.
Anthropic said it received the government's order at 5:21 p.m. ET (2121 GMT) Friday and noted that the letter did not spell out the specific security concern in detail.
According to the company, the government believes an individual has discovered a method to "jailbreak" — or sidestep — the safety controls built into Fable 5. Anthropic countered that the vulnerabilities cited were already on its radar, were "minor" in nature, and were replicable using other freely available AI systems without requiring any special bypass technique.
The company argued that it had invested thousands of hours stress-testing Fable 5's safety architecture before its launch, working alongside the US government, the UK's AI safety regulator, and other bodies — none of which found a complete method to circumvent its protections.
"We suspect that perfect jailbreak resistance is not currently possible for any model provider," the company said, cautioning that all AI safety systems carry some degree of vulnerability and that highly effective bypass techniques will likely surface for any model at some point.
Anthropic disclosed that the government had furnished only verbal details of a single, narrow bypass method, which it characterized as "asking the model to read a specific codebase and fix any software flaws." It also said that after reviewing what it believes to be the report underpinning the order, it found the same capability to be "widely available from other models (including OpenAI's GPT-5.5), and is used every day by the defenders who keep systems safe."
While pledging to comply with the directive, Anthropic maintained that a contained issue of this kind should not justify withdrawing a commercial AI model relied upon by hundreds of millions of users, warning that applying such a standard industry-wide "would essentially halt all new model deployments for all frontier model providers." The company described the situation as the product of "a misunderstanding" and said it is working urgently to restore access.
Legal Disclaimer:
EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
